Efficient initiation of adenovirus DNA replication requires the presence of specific terminal nucleotide sequences that collectively constitute the viral origin of replication. Using plasmids with deletions or base substitutions in a cloned segment of DNA derived from the terminus of the adenovirus 2 genome, we have demonstrated that the origin contains two functionally distinct regions. The first 18 bp of the viral genome are sufficient to support a limited degree of initiation. However, the presence of a sequence in the region between nucleotides 19 and 67 greatly enhances the efficiency of the initiation reaction. This region contains a specific binding site for a protein present in uninfected cells (KD = 2 X 10(-11) M). The bound protein protects the DNA segment between base pairs 19 and 43 from attack by DNAase I. Studies with deletion mutants indicate that binding of the cellular protein is responsible for the enhancement of initiation.

The adenovirus origin of DNA replication contains three functionally distinct sequence domains (A, B, and C) that are essential for initiation of DNA synthesis. Previous studies have shown that domain B contains the recognition site for nuclear factor I (NF-I), a cellular protein that is required for optimal initiation. In the studies reported here, we used highly purified NF-I, prepared by DNA recognition site affinity chromatography (P. J. Rosenfeld and T. J. Kelly, Jr., J. Biol. Chem. 261:1398-1408, 1986), to investigate the cellular protein requirements for initiation of viral DNA replication. Our data demonstrate that while NF-I is essential for efficient initiation in vitro, other cellular factors are required as well. A fraction derived from HeLa cell nuclear extract (BR-FT fraction) was shown to contain all the additional cellular proteins required for the complete reconstitution of the initiation reaction. Analysis of this complementing fraction by a gel electrophoresis DNA-binding assay revealed the presence of two site-specific DNA-binding proteins, ORP-A and ORP-C, that recognized sequences in domains A and C, respectively, of the viral origin. Both proteins were purified by DNA recognition site affinity chromatography, and the boundaries of their binding sites were defined by DNase I footprint analysis. Additional chracterization of the recognition sequences of ORP-A, NF-I and ORP-C was accomplished by determining the affinity of the proteins for viral origins containing deletion and base substitution mutations. ORP-C recognized a sequence between nucleotides 41 and 51 of the adenovirus genome, and analysis of mutant origins indicated that efficient initiation of replication is dependent on the presence of a high-affinity ORP-C-binding site. The ORP-A recognition site was localized to the first 12 base pairs of the viral genome within the minimal origin of replication. These data provide evidence that the initiation of adenovirus DNA replication involves multiple protein-DNA interactions at the origin.Initiation of adenovirus DNA replication takes place at either terminus of the linear viral genome (for a review, see reference 23). The initiation reaction involves the formation of a covalent linkage between dCMP, the first nucleotide of the nascent DNA chain, and a virus-encoded primer protein, referred to as the 80-kilodalton (kDa) preterminal protein (pTP) (7,9,28,36,45). The protein-nucleotide bond formed during this reaction has been identified as an ester linkage between the c-phosphoryl group of dCMP and the 3-OH of a serine residue in the pIP (11). The synthesis of a new viral DNA strand takes place by extension from the free 3' hydroxyl group present in the pTP-dCMP initiation complex.Analysis of the replication of mutant viral genomes in vitro has demonstrated that the initiation of viral DNA replication is dependent on the presence of specific nucleotide sequence domains within the terminal region of the viral DNA (10, 12, 21, 25, 26, 38. 45-47. 49 nucleotides 40 and 51 (domain C) increas...

The initiation of adenovirus DNA takes place at the termini of the viral genome and requires the presence of specific nucleotide sequence elements. To define the sequence organization of the viral origin, we tested a large number of deletion, insertion, and base substitution mutants for their ability to support initiation and replication in vitro. The data demonstrate that the origin consists of at least three functionally distinct domains, A, B, and C. Domain A (nucleotides 1 to 18) contains the minimal sequence sufficient for origin function. Domains B (nucleotides 19 to 40) and C (nucleotides 41 to 51) contain accessory sequences that significantly increase the activity of the minimal origin. The presence of domain B increases the efficiency of initiation by more than 10-fold in vitro, and the presence of domains B and C increases the efficiency of initiation by more than 30-fold. Mutations that alter the distance between the minimal origin and the accessory domains by one or two base pairs dramatically decrease initiation efficiency. This critical spacing requirement suggests that there are specific interactions between the factors that recognize the two regions.

The Drosophilu epidermal growth factor receptor homolog (DER) displays sequence similarity to both the epidermal growth factor (EGF) receptor and the neu vertebrate proteins. We have examined the possibility of deregulating the tyrosine kinase activity of DER by introducing structural changes which mimic the oncogenic alterations in the vertebrate counterparts. Substitution of valine by glutamic acid in the transmembrane domain, in a position analogous to the oncogenic mutation in the rat neu gene, elevated the in vivo kinase activity of DER in Drosophilu Schneider cells sevenfold. A chimera containing the oncogenic neu extracellular and transmembrane domains and the DER kinase region, also showed a threefold elevated activity relative to a similar chimera with normal neu sequences. Double truncation of DER in the extracellular and cytoplasmic domains, mimicking the deletions in the v-erhB oncogene, did not however result in stimulation of in vivo kinase activity. The chimeric constructs were also expressed in monkey COS cells, and similar results were obtained. The ability to enhance the DER kinase activity by a specific structural modification of the transmembrane domain demonstrates the universality of this activation mechanism and strengthens the notion that this domain is intimately involved in signal transduction. These results also support the inclusion of DER within the tyrosine-kinase receptor family.Transmembrane receptors constitute a significant fraction of the diverse family of tyrosine kinases (reviewed in [l]). Many of them have been implicated in transformation processes and identified as potent oncogenes. Structural changes in these receptors have been associated with increased kinase activity which in turn gives rise to oncogenic potential. In some cases, major structural changes lead to activation, while in others, a single amino acid substitution is sufficient. The mechanisms by which structural changes lead to deregulated kinase activity provide a key to understanding the normal mechanism of signal transduction by these receptors.Major truncations appear to represent one class of activating mechanisms. The v-erbB oncogene has retained only 65 amino acids at the extracellular domain, and lost the C-terminal32 or 44 residues [2]. Similarly, avian leukosis virus insertions into the c-erbB gene give rise to a protein with the same extracellular deletion [3]. Comparison of the v-ros and c-ros sequences demonstrates that v-ros has lost most of the normal extracellular region [4]. Finally the v-kit oncogene, which contains no transmembrane or extracellular domains is derived from a cellular gene coding for a putative transmemCorrespondence to B.-2. Shilo, Department of Molecular Genetics and Virology, Weizmann Institute of Science, IL-76100 Rehovot, Israel Abbreviations. EGF, epidermal growth factor; DER, Drosophilu EGF receptor homolog; SV40, simian virus 40; WT, wild-type DER protein; KD, kinase deletion; DT, double truncation; TM, transmembrane point mutation; KM, kinase inactivation point mutat...

RNA synthesis in eucaryotes takes place on template molecules that are activated by stably associating with limiting transcription factors. In this paper we demonstrate that such stable transcription complexes can be specifically sedimented from in vitro transcription reaction mixtures by mild centrifugation. This occurs with stable complexes of genes transcribed by all three classes of eucaryotic RNA polymerase and with S-100 as well as whole-cell extracts. However, the transcriptional capacity of the isolated complex differs for the three polymerase classes. The pelleted ribosomal DNA (polymerase I) complex contains all the factors necessary for transcription, each purified 25-to 50-fold, whereas the pelleted adenovirus major late promoter (polymerase II) complex lacks a factor that remains in the supernatant. In the case of 5S DNA (polymerase III), a necessary factor associates slowly with the sedimentable complex. Notably, the interactions responsible for this rapid sedimentation are specific for DNA molecules in stable complexes, suggesting that the in vitro sedimentable complex mirrors the in vivo structural organization of active genes.Eucaryotic transcription factors stably associate with the DNA molecules that serve as templates for RNA synthesis. Formation of such stable transcription complexes has been demonstrated in vitro with genes transcribed by all three classes of polymerase (1, 3, 5-7, 9, 11, 22, 24, 25, 27) and thus appears to be the general means of maintaining the activated state of eucaryotic genes. The constituents of these stable complexes and their functional roles are currently under considerable investigation. The DNA sequences involved in stably binding polymerase I, II, and III transcription factors have been delineated for genes that encode 45S rRNA (B. Sollner-Webb, J. Tower, V. Culotta, and J. Windle, in J. Setlow and A. Hollaender, ed., Genetic Engineering, in press), the adenovirus major late transcript (7), and 5S RNA (21), respectively. In all cases, stable complex formation utilizes the same template region that was previously identified as the promoter. Considerable progress has also been made in isolating and characterizing the proteins involved in the transcription complex. In many cases, however, these studies have been hindered by the difficulty in purifying these factors without effecting significant losses of activity; despite much effort, most eucaryotic transcription factors have only been purified to a limited extent.As an alternative to traditional chromatographic methods, we have utilized a biological fractionation as the first step in purifying transcription factors, taking advantage of their specific binding to template molecules. We show that stable complexes and the associated transcription factors can be quantitatively and selectively sedimented from in vitro transcription reaction-inixtures, whereas the bulk of the extract protein remains soluble. This procedure is effective in isolating transcription complexes of all three classes of eucaryotic polymerase. ...